We conducted a six-month, randomized, double-blind trial of lecithin therapy in early-onset Alzheimer's disease. We hypothesized that such therapy would retard the progression of the clinical and neuropsychological manifestations of this illness. Of the 73 referred patients, 37 met strict requirements for diagnosis and compliance. The 21 placebo and 16 lecithin-treated patients (mean age 63 years) had a comparable degree of severity of dementia (mean Clinical Dementia Rating 1.6). Lecithin therapy produced an increase in mean plasma choline levels from a baseline of 15.9 to 28.8 nmol/ml. Patients were evaluated by the physician using clinical assessments (CDR, Lawton ADL and other rating scales) and by the neuropsychologist who determined the outcome of therapy on a battery of tests (Mini Mental State Examination, Wepman Aphasia Screen, Verbal Fluency Test, Verbal Selective Reminding Test and Spatial Memory Test). Only 6 (37.5%) of the 16 lecithin-treated patients were considered by the neurologist to be clinically stable or improved as compared to 12 (57.1%) of the 21 patients given placebo (difference -19.6%, 95% confidence limits of -51% to 12%). The neuropsychologic scores showed no differences in the stability of the dementing process over time between the lecithin-treated (50.0%) and placebo (47.6%) groups. On the basis of these clinical and neuropsychological findings, it appears that lecithin alone has no important therapeutic effect in early-onset Alzheimer's disease.